Article Details
2026-11-23BXT Data Editorial

SCAYLE ZEOS marketplace B2B 2026

SCAYLE ZEOS marketplace B2B 2026 article image

"Customers will never have to leave the trusted Zalando app" — that line in the company's own release explains more than the headline growth numbers. Through November 3, 2026, the platform reported 7.5 percent GMV growth to 4.5 billion euros and confirmed Zalando Extended, an AI agent that will check out products from other online shops inside the app from 2027. What changed this quarter was not the discount or the delivery speed but who is allowed to press buy. For ecommerce operators, the marketplace is moving from selling its own catalogue to brokering someone else's, with an agent in charge of the transaction.

The quarter in numbers

Known factsUnverifiedNeeds verification
The reported GMV growth, delivery coverage and fee figures come from the named sources.Whether unit economics hold at full rollout is not disclosed.Check each figure against the primary report before quoting it.
Dates and named partners are taken from the cited coverage.Store-level availability and price parity are not published in aggregate.Verify locally with a sample basket in the relevant market.

Zalando's official release, published via EQS News, reports group GMV up 7.5 percent to 4.5 billion euros and revenue up 5.1 percent to 3.2 billion euros. Adjusted EBIT rose 10.6 percent to 107 million euros, with the margin expanding 0.2 percentage points to 3.4 percent. Active customers reached a record 63.2 million, up 2.9 percent, and the business-to-business unit was the standout: B2B revenue grew almost 28 percent to 354 million euros, with adjusted EBIT more than doubling to 42 million euros.

The platform business is gaining weight. Partner business and retail media grew at a double-digit rate, and the partner share of B2C GMV rose 2.9 percentage points to 33.6 percent, edging toward the 2028 target of 40 to 50 percent. Coverage of the results notes that retail-media revenue grew nearly 18 percent and average spend per customer rose 2.9 percent to 308 euros. Those are the foundations on which an agent-led shopping model can be built.

How the AI agent actually works

Zalando Extended adds products from other online shops — items Zalando does not sell itself — to its platform. Rather than sending the shopper to another website, Zalando's own AI agent securely completes the purchase inside the app. The company plans to start with complex categories such as bulky home goods, with Westwing as the first pilot partner; from 2027 the agent will be able to check out Westwing products directly.

For partner shops, the appeal is reach without friction. Because the agent handles the transaction, retailers can access Zalando's customer base without complex technical integration or changes to their own logistics. They can also advertise through Zalando's retail-media business. The catch is that the product page becomes an input to an algorithm, not just a page a human browses. Product data quality, clear attributes and reliable availability become prerequisites for being selected.

Why agent-led discovery changes the category game

The conversational Zalando Assistant has already been used by more than 22 million unique users this year. Industry analysis places Extended within a broader move toward agent-led shopping, in which AI systems research and buy on a shopper's behalf. When an agent shortlists products, the considerations are different from a human scrolling a results page: structured attributes, price consistency and stock reliability outweigh brand storytelling.

That shifts where category growth comes from. A brand that wins on compelling imagery but poor structured data may be invisible to the agent. Conversely, a product with clean attributes, consistent pricing and dependable availability can be surfaced repeatedly without the shopper ever running a conventional search. This is a new kind of product listing discipline.

The margin logic behind the platform pivot

Zalando confirmed full-year guidance of 12 to 17 percent GMV and revenue growth in the lower half of the range, and adjusted EBIT of 680 to 720 million euros. The B2B unit, built around logistics service ZEOS, marketplace integrator Tradebyte and ecommerce software provider SCAYLE, supplies the infrastructure that makes a platform model economically viable. SCAYLE expanded enterprise deals in the US and UK, while ZEOS secured customers including Lascana and Champion.

In other words, the AI agent is the consumer-facing edge of a platform strategy whose profit engine sits in B2B services and retail media. Treating Extended as a standalone gadget would miss where the value accrues. The company confirmed full-year guidance of 12 to 17 percent GMV and revenue growth in the lower half of the range, with adjusted EBIT of 680 to 720 million euros. That outlook rests on logistics service ZEOS, marketplace integrator Tradebyte and ecommerce software provider SCAYLE, whose enterprise deals in the US and UK supply the infrastructure a platform model needs. For consumer-goods teams, the implication is that product innovation now has to account for how an agent reads a new product: clear attributes, verified consumer need and consistent packaging data are part of being discoverable before a human ever browses.

What operators should do now

Operator taskWhat to checkFailure signal
Product data auditSize, material, dimensions, compatibility fields completeProduct skipped by the agent shortlist
Channel alignmentPrice and availability consistent across channelsAgent switches to another seller
Retail-media budgetVisibility spend treated as discovery costListing invisible in agent-led results

First, audit product data. If an agent is choosing products, attributes such as size, material, dimensions and compatibility must be complete and consistent across the catalogue. Second, align price and availability across channels, because an agent that encounters a mismatch will simply pick another seller. Third, treat retail media as part of the cost of discovery rather than an optional add-on.

The ecommerce shift toward agents does not make branding irrelevant, but it changes the sequence. Before a shopper forms a preference, the agent has already filtered the field. Brands that invest in clean, verifiable product information will be in that field; those that do not will be filtered out before a human ever sees them.

  1. Step 1 — Merchandising lead: pull the current SKU list for the named stores and record availability (in stock / out of stock) plus on-shelf price for the tracked items.
  2. Step 2 — Channel manager: open the delivery apps in that market, rebuild the same basket, and capture the delivered price after coupons and fees.
  3. Step 3 — Analyst: compare store and app prices, flag items where distribution coverage or availability breaks, and log the sampling date.
fieldexample
store_idPL-WAW-0142
skuMILK-1L-BRAND-A
availabilityin stock
on_shelf_price4.29 PLN
app_delivered_price5.10 PLN
sample_date2026-11-18

FAQ

  • What is Zalando Extended in practical terms?

    An AI agent that buys products from other online shops inside the Zalando app, so the shopper never leaves the platform. It starts with complex home-goods categories and goes live more broadly in 2027.

  • How does agent-led shopping change product listing requirements?

    Attributes, price consistency and availability matter more. An agent selects from structured data, so incomplete or inconsistent product pages can be skipped before a shopper sees them.

  • Is the AI agent the main profit driver?

    No. The margin comes from B2B growth and retail media. Extended is the consumer edge of a platform model whose economics sit in logistics, marketplace software and advertising.

Sources

Recommended